This grant application describes a cell-based assay for detecting molecular recognition and catalysis that can be used to evolve proteins with new functions. There is tremendous interest in being able to engineer proteins with new specificities and new activities for use as reagents for biomedical research, diagnostics and therapeutics for the health care community, and tools for the pharmaceutical industry. The screen builds from existing technology for dimerizing proteins inside a cell with dimeric ligands via the ligands' receptors (CIDs). By replacing one of the ligand-receptor pairs with potential binding partners, binding can be detected. By replacing the chemical linker between the two ligands with a bond and adding an enzyme, the assay can be used as a read-out for bond formation or bond cleavage. In Preliminary Results dexamethasone-methotrexate CIDs with non-cleavable and cleavable linkers have been developed. Aim 1 outlines our plans to evolve a protein receptor for estradiol that can be used in medical diagnostics for monitoring estrogen levels in women. We have developed a docking algorithm to pick several monomeric proteins from the PDB as the starting protein scaffolds. We plan to mutagenize these proteins using existing methods and then select for high affinity, specific receptors by screening for binding to estradiol and against binding to other common steroids. Aim 2 describes our plans to modify the yeast two-hybrid assay to detect catalysis and then evolve a penicillin-binding protein into a cephalosporinase enzyme. Penicillin-binding proteins are the target of penicillin antibiotics and are believed to be the evolutionary precursors of cephalosporinases, the bacterial resistance enzymes that hydrolyze and inactivate these antibiotics. Because of the evolutionary relationship, the PBPs present a tractable first target for enzyme evolution. Moreover, this project should provide insight into how bacteria evolve antibiotic resistance and the mechanism by which the resistance enzymes hydrolyze the antibiotic. Finally, in Aim 3, we propose to develop a bacterial CID so that future protein evolution experiments can be carried out in bacteria. Bacteria have faster doubling times and higher transformation efficiencies than yeast, and so a bacterial CID system should facilitate the evolution experiments.